@Kris Bock Thank you for your answer. Is it possible to also register the output as a data asset? I tried to construct the Output object with the "name" parameter (as it works in normal ml pipelines) but here it doesn't seem to have any effect. Is there another way or did I miss something?
Azure ML Batch Endpoint: Configuring the output
Hello,
I created a batch endpoint in Machine Learning Workspace and I would like to configure the output path. I follow the tutorial https://zcusa.951200.xyz/en-us/azure/machine-learning/how-to-deploy-model-custom-output?view=azureml-api-2&tabs=python and under the part Creating a scoring script, it shows how the environment variable AZUREML_BI_OUTPUT_PATH is used to get the output bath. And in the part Creating the deployment, output_action is set to BatchDeploymentOutputAction.SUMMARY_ONLY which means that the user script will store the output. I run a sample job with this configuration and the results is written somewhere in workspaceblobstore. But I would like to write the results to another Storage Account. My question can I write output to another blob storage or Data Lake Storage? If so how? If this is not possible, can I somehow overwrite the environment variable AZUREML_BI_OUTPUT_PATH so that I can write to a specific place that I want in workspaceblobstore?
Thank you for your time and help.